Google AI Platform, TensorFlow, Microsoft Azure, Rainbird, Infosys Nia, Wipro HOLMES, Dialogflow, Premonition, Ayasdi, MindMeld, Meya, KAI, Vital A.I, Wit, Receptiviti, Watson Studio, Lumiata, and Infrrd. These are just some of the top artificial intelligence (AI) platforms today.
Definitions of what a digital platform is made up suggest it is different things to different people. A common observable characteristic is that digital platforms serve as a medium by which transactions or services are performed.
What is certain is that the hardware, software, and connectivity elements that make up digital platforms will vary depending on the application or use.
FutureCIO spoke to Louis LW Teoh, regional director of commercial operations for APJ for Genesys, for his take on what digital platforms are in the context of AI and data analytics delivery and consumption.
Digital platforms
Teoh defines digital platforms as applications built upon computational infrastructures that often work collectively within an ecosystem of applications – in recent years initiatives have often flown under the banner of “digitisation”. He opines that motivations are usually associated with business growth and efficiency.
Using the analogy of how telcos have moved into software-defined networks space; he suggested that the use of digital platforms has done the same to streamline & transform businesses. He cited the move from wet signatures to e-signatures as another example.
“CCaaS and CX have taken a similar digitisation trajectory. This is critical in expanding or pivoting the business into new business models, leveraging the new capabilities from the digital platforms,” he added.
Optimisation challenges
It is stipulated that for AI to be effective it needs a large amount of data. OpenAI’s GPT-4, for example, was trained on 570GB of datasets including web pages, books, and other sources. Its predecessor, GPT-3, reportedly cost US$12 million for a single training run.
Teoh points out that data analytics and AI systems consume data generated across digital platforms to glean insights, support decision-making, and automate processes for optimal outcomes.
He cautioned that without an intentional strategy and approach to optimize one’s digital platforms, AI and D&A will be impeded & limp along.
He further posits that over the years, organisations have started accruing “Technical Debt” – caused in part by independent system owners and platform owners standing up digital platforms in isolation, where consideration for AI & D&A is an after-thought.
Best practices for optimising digital platforms
Data Collection and Instrumentation by Design. Teoh opines that as digital platforms are being implemented, a part of the man-in-the-middle is often & inadvertently removed from the processes.
“The consideration here is to ensure that data which would have otherwise been captured by the human process steps continues to be captured – this requires intentional effort to retain the resolution of the data fabric. Doing this well allows layering in new forms of data now possible through the structured approach of data collection via digital systems, which may be inconsistent before.”
“For AI & D&A, we need to know as much as possible upfront about how we intend to analyse and leverage the data – to shape the design of this,” he suggested.
Don’t miss the Forest for the Trees. Teoh says that the immediate focus for most organisations should be on unlocking growth and efficiency opportunities instead of burning precious cycles in figuring out grandiose, pivotal transformations.
“New digital standards continue to emerge & evolve, even as organisations re-engineer their business model and processes. At this juncture, when approaching the perennial question of build-versus-buy, often the first questions should be whether the choice of either brings competitive advantage, coupled with time-to-value, that is viable to the business,” he says.
He says it is crucial to consider the key considerations of clients and partners such as level of platform homogeneity, solution interoperability, flexibility of direct & system integration services engagements, time-to-results, ease of seamless upgrade paths, and platform integrity.
Simplification. “When seeking to build AI capabilities, ensure talent and projects are not overly fixated on megalithic enterprise digital transformations without considering the full spectrum of wins narrow to broad AI can bring – there is intrinsic value to be yielded across the spectrum – doing this well along with change management can bring about incredible institutionalized competitive edge,” he says.
He adds that it is important to consider the intrinsic value of fit-for-purpose UX, global/regional presence, as well as openness and interoperability with one’s enterprise AI & applications ecosystem.
He also suggests organisations consult with credible peer groups & analyst viewpoints to gauge any given vendor’s vision and, as critically, one’s ability to execute.
“Thoughtfulness to ensure the ease of end-user adoption, the pace of innovation and geographic availability & resilience are important aspects, beyond simply the near-term needs of procuring a set of features & functions.“
Securing and maintaining compliance
Asked how organisations can ensure that their digital platforms are secure and compliant with relevant regulations and standards, Teoh suggests beginning by finding the right balance when considering proprietary applications.
“Strive for the right mix of proprietary application vs non-proprietary ones, that is appropriate to the level of organisational maturity while meeting business objectives,” he added.
For Teoh, finding the right vendors is also important in securing and maintaining compliance. He says it is vital to evaluate the completeness of vision, strengths in execution, and stability & consistency in its growth trajectory when seeking vendors.
He says that a sizable & proven customer base shows the breadth and depth of a vendor’s market presence.
A strong partner network as well as established is important to ensure talent availability and tech stack viability.
Louis LW Teoh
Preparing for the future of digital platforms for AI and D&A delivery and consumption
Teoh posits that the ability to harness digital platforms is not just technical but should also focus on culture and process.
Human-augmented AI. “ Generally, the basis of “AI-augmented” work focuses on increasing efficiency, while “human-augmented” work focuses on increasing effectiveness.
Organisations and workers may find it assuring to know that some point of equilibrium with the right blend of AI & human augmented is required,” he says.
He believes that AI can augment skilled workers as they develop critical thinking skills and higher-level cognitive abilities.
For this to happen, Teo says this means re-calibrating organisational workflow, re-designing roles & hiring for the future, and even overhauling organisational workflows and exploring new operating models.
“Some more radical, some less. The ability for organisations to optimize their workforce wisely in placing people & technology assets,” he says.
Purpose, People, Programs, Provisions. Teoh posits that it is important for executives to initiate their organisation’s clarity of purpose which paves the way for identifying people/talent, programs, and provisioning.
Starting with the end in mind. “Few AI and D&A projects get to see the light of deployment into production. The ability to sus out viable initiatives that can be operationalized, from distractions, is a prized ability that can save organisations time & effort,” he says.
Teoh points out that what fuels deeper investment and fosters organisational confidence are the organisation’s success and quantifiable results.
